The 60-minute AI readiness self-assessment

Thirty questions, five sections, no jargon. Score yourself honestly and you’ll have a defensible read on whether your business is actually ready to invest in AI - or whether you’d be lighting money on fire. Built from the diligence we run before scoping any client engagement.

Section 1 - Data readiness (max 12)

1. We can find any customer record (name, contact, history) within 60 seconds.
2. Our data lives in named systems (CRM, accounting, ops tool) - not in someone’s head or in 47 spreadsheets.
3. We have written process docs for at least one repeating workflow.
4. We could export our key business data to a CSV today if asked.
5. We know what data we hold that counts as personal data under UK GDPR.
6. We have a backup of our key systems from the last 7 days.

Section 2 - Process maturity (max 12)

1. At least one core process (quoting, invoicing, onboarding) runs the same way every time, regardless of who’s on shift.
2. We can name the top 3 tasks that eat the most owner / manager time each week.
3. We measure something operational - jobs per week, response time, conversion rate, anything.
4. When something goes wrong, we know where to look (logs, tickets, email trails) without ringing five people.
5. We have a tool for the work we do - not just email and a shared drive.
6. New hires can do the basics within a fortnight without shadowing for a month.

Section 3 - Team capability (max 12)

1. Someone in the team has used ChatGPT, Claude or similar for a real task at least weekly.
2. We have at least one person comfortable installing a new piece of software and getting it talking to existing tools.
3. We’re willing to spend 1-2 hours per person on training before deploying anything new.
4. Someone in the team can read a contract, an API docs page, or an invoice from a vendor critically.
5. We’ve adopted a new tool successfully in the last 12 months.
6. We have a working understanding of our 2-3 main software subscriptions and what they cost.

Section 4 - Risk posture (max 12)

1. We have a written policy (even one page) on what staff should and shouldn’t paste into AI tools.
2. We could explain to a client what data we collect on them and how we store it.
3. We have a way of approving / rejecting AI-generated output before it leaves the building.
4. We know which of our work is regulated (financial advice, legal, healthcare etc) and what that means.
5. We’re happy to put our names to AI-assisted work publicly, with appropriate disclosure.
6. We have an answer for the question “what if the AI gets it wrong?” that doesn’t involve hoping it doesn’t.

Section 5 - Use-case fit (max 12)

1. We’ve identified one specific repeating task we’d like AI to help with - not “use AI somehow”.
2. That task is well-defined enough that we could write step-by-step instructions for a new hire.
3. The task happens at least weekly (so any improvement compounds).
4. A wrong answer on that task is recoverable, not catastrophic.
5. We could measure whether the AI is doing better, worse, or the same as today.
6. We have a budget in mind for trying this - even if it’s small.

Read your score

Most businesses we score for the first time land in the 21-35 band. That’s normal. It’s also the point where getting one bounded pilot right pays for itself within a quarter.

Common low-score patterns and how to fix them

• Low on Data, high elsewhere.The team’s sharp but the systems are scattered. Pick one system to be the source of truth (usually the CRM or the accounting tool) and move everything material into it before you bolt AI on top.
• Low on Process, high on Capability. A clever team firefighting. Document one process this week. AI multiplies whatever process you point it at - if the process is chaos, you get faster chaos.
• Low on Risk posture only. You have one weekend of admin to do: write a 1-page AI policy, decide who reviews AI output, agree what counts as disclosable to clients. Then carry on.
• High on everything except Use-case fit.Resist generic AI tooling. Talk to two or three people who actually do the work and ask what they’d stop doing tomorrow if they could. That’s your pilot.

Want a second opinion on your score?

Send us your totals and the use case you have in mind and we’ll send back a one-page read on whether it’s worth pursuing - free, no follow-up obligation. Use the contact form or email enquiries@glassworkanalytics.com.